In the post The Munger Games, inspired by my first-ever attendance at the Berkshire annual meeting and purchase and reading of Poor Charlie’s Almanack, I promised more commentary on Charlie Munger’s book once I’d reflected on it.
Interesting, I can see the airline analogy, given the amount of expenditures and what the return of capital it may or may not bring.
I watch a lot of people becoming dependent on AI. If they don’t know the answer, instead of taking time to think, they will immediately and many times carelessly dump information into it.
Off topic, but how does this enhance the human brain long term?
I know one thing: if people continue to use free AI to ask trivial questions like mine about Steely Dan, or create images like "Taylor Swift as a hillbilly", it will have to become far more energy efficient. The demands on the grid can't be justified by foolish uses. Either that or AI will become mainly a B2B thing, with B2C only available via subscription or one-off fees.
It feels almost silly to say that AI is "undifferentiable" similar to Airplanes or Mobile Phones, given the nature of Generative AI. If you ask it the same prompt 1000 times, you get 1000 answers! But, I can't disagree that's where it's looking to go, but not because of a fundamental quality of AI, but because of the diminishing gains.
The difference between an AI as smart as a 5 year old is miles worse than an AI as smart as a 15 year old. The difference, though, to a 25 year old is less. And to a 35 year old? Forget about it.
So even if I made an AI with greater command on the English language, it's visibility would be marginal. Think of it less like Cereal and more like Sodas, Pepsi and Coke are undoubtedly different, but how much? Not enough to care.
But this presents a great opportunity. What made Coke so successful? Distribution and Marketing. They had vending machines everywhere, so people reached for Coke, even if it wasn't that much better than Pepsi. Then they marketed like hell, so they'd think of Coke 10 times before Pepsi.
ChatGPT has that market reach and distribution right now, even if DeepSeek or Claude make big splashes, I feel OpenAI has a brand awareness that money simply can't buy, and they're likely going to keep a stronghold on the consumer side.
The B2B side is where things get interesting. Businesses will be much more attuned to differences, however minute, and they'll have more vocal and complex demands, which AI labs can optimize for.
In terms of investments, I think an underserved market will be going for differentiation (especially at this age where AI is as worse as it'll ever be). But it'll be risky long-term because the labs that focused on raw processing will eventually beat out a hyper-optimized weaker model.
Cereals are daily staple so we cannot compare that to Airlines, which are slightly occasional unless business travel a du jour. AI will fit the cereal class just like iPhone Apps..
Interesting, I can see the airline analogy, given the amount of expenditures and what the return of capital it may or may not bring.
I watch a lot of people becoming dependent on AI. If they don’t know the answer, instead of taking time to think, they will immediately and many times carelessly dump information into it.
Off topic, but how does this enhance the human brain long term?
I know one thing: if people continue to use free AI to ask trivial questions like mine about Steely Dan, or create images like "Taylor Swift as a hillbilly", it will have to become far more energy efficient. The demands on the grid can't be justified by foolish uses. Either that or AI will become mainly a B2B thing, with B2C only available via subscription or one-off fees.
It feels almost silly to say that AI is "undifferentiable" similar to Airplanes or Mobile Phones, given the nature of Generative AI. If you ask it the same prompt 1000 times, you get 1000 answers! But, I can't disagree that's where it's looking to go, but not because of a fundamental quality of AI, but because of the diminishing gains.
The difference between an AI as smart as a 5 year old is miles worse than an AI as smart as a 15 year old. The difference, though, to a 25 year old is less. And to a 35 year old? Forget about it.
So even if I made an AI with greater command on the English language, it's visibility would be marginal. Think of it less like Cereal and more like Sodas, Pepsi and Coke are undoubtedly different, but how much? Not enough to care.
But this presents a great opportunity. What made Coke so successful? Distribution and Marketing. They had vending machines everywhere, so people reached for Coke, even if it wasn't that much better than Pepsi. Then they marketed like hell, so they'd think of Coke 10 times before Pepsi.
ChatGPT has that market reach and distribution right now, even if DeepSeek or Claude make big splashes, I feel OpenAI has a brand awareness that money simply can't buy, and they're likely going to keep a stronghold on the consumer side.
The B2B side is where things get interesting. Businesses will be much more attuned to differences, however minute, and they'll have more vocal and complex demands, which AI labs can optimize for.
In terms of investments, I think an underserved market will be going for differentiation (especially at this age where AI is as worse as it'll ever be). But it'll be risky long-term because the labs that focused on raw processing will eventually beat out a hyper-optimized weaker model.
Cereals are daily staple so we cannot compare that to Airlines, which are slightly occasional unless business travel a du jour. AI will fit the cereal class just like iPhone Apps..
No offense but I think you’re missing the point.
I don’t think so, FWIW you may want to relook at the sample population of AI enterprises you are using and extrapolating at large.